Developer Experience Engineer
A real software engineer whose users are other engineers — and at scale, one of the best-paid seats in the building.
- Entry
- $110k
- Mid
- $165k
- Senior
- $230k+
- Demand
- Rising
Also called Developer Productivity or Platform Tooling, this is the team that builds the build system, the internal CLI, the CI/CD pipeline, and the golden paths every other engineer depends on. The work is invisible from the outside — there's no app to show your friends — but the leverage is enormous: shave ten minutes off a build and you've handed time back to a thousand engineers at once. At large companies it's both prestigious and very well paid.
The myth
It's just babysitting CI — not real product work.
The reality
You own the build system, the test infrastructure, the internal platform and the tooling thousands of engineers can't ship without. It's deep systems engineering with the highest leverage in the org.
cat ./what_you_actually_do.md
- Own and speed up CI/CD and the build system — every minute you cut is multiplied across the whole org.
- Build internal CLIs, scaffolding, and 'golden path' templates so engineers fall into the pit of success by default.
- Hunt down and kill flaky tests, slow steps, and the daily toil that quietly drains engineering hours.
- Treat developer productivity as a measurable product: instrument the inner loop, find the bottleneck, fix it.
- Run big, scary migrations (build tooling, monorepo changes) without grinding everyone else to a halt.
cat ./why_underrated.md
Nobody markets it because there's no logo to point at — the users are internal and the wins show up as an absence of pain, not a launch. So students never hear 'developer productivity' framed as a career, even though every large engineering org has a whole team doing it. The people who find it tend to love it: it's pure engineering, the leverage is addictive, and the comp matches senior product engineering while the work is often calmer than feature delivery. It's quietly one of the smartest places for a strong generalist engineer to land.
grep -i 'good fit' ./who.md
- Engineers who get more satisfaction from unblocking ten people than shipping one feature.
- Systems thinkers who like infrastructure, tooling, and build internals.
- People allergic to repeated manual toil — the kind who automate it on reflex.
cat ./pay.md
Comp tracks senior software engineering because it is software engineering — your customers just happen to sit two desks over. At big tech, Developer Productivity is a respected discipline where total compensation clears $250k for senior engineers, since the multiplier on everyone else's output is impossible to ignore.
./break_in.sh
Be the fixer on your own team
Speed up the slow build, fix the flaky test, write the script everyone keeps asking for. That reputation is the whole interview.
Go deep on one build/CI system
Learn a real build tool (Bazel, Gradle) and a CI system end to end. Depth here is rarer than it sounds and instantly valuable.
Contribute to dev tooling in the open
Open-source build tools, CLIs, and CI actions are always short on contributors — visible, attributable experience that maps straight to the job.
Target orgs big enough to have the team
Companies past a few hundred engineers almost always have a Platform / DevProd group. That's where the role (and the comp) actually exists.
tail -f ./a_day.log
- 09:00CI got 20% slower this week; profile the pipeline and trace it to an un-cached dependency step.
- 11:00Ship a new internal CLI command that scaffolds a service with logging, CI, and deploy wired in.
- 14:00Quarantine a flaky test that's been eroding everyone's trust in the suite, and fix the root cause.
- 16:00Review a build-system migration plan, focused on not breaking a thousand people's afternoon.
ls ./toolbelt
- A systems language (Go)
- Bash & Python
- CI/CD systems
- Bazel / build systems
- Docker
- Pipeline observability